DEVELOPMENT OF A METAHEURISTIC-BASED LOAD BALANCING ALGORITHM TO MITIGATE OVERLOADING IN FEDERATED CLOUD INFRASTRUCTURES

In a subscription-based service known as cloud computing, clients have scheduled access to shared resources such as data, software, and other assets as needed. Despite, several benefits, cloud computing, still faces significant difficulties. Load balancing which is the capacity of the cloud infrastructure to equally distribute tasks resources in the cloud environment has significant issues. A new idea of cloud deployment referred to as cloud federation was started in order to offer solutions to the issue of load unbalancing in the cloud infrastructures. However, in a federated cloud system, transparent workload sharing among participating Cloud Service Providers (CSP) is challenging. This research work presents the development of a load balancing algorithm in a simulated federated cloud environment by considering inter-cloud and intra-cloud loads. The inter-cloud load balancing was realized using Ant Colony Optimization (ACO) algorithm as the Federated Cloud Load Balancer (FedCloudBalancer) while the intra-cloud aspect was realized with an existing throttled load balancing algorithm. The implementation of the FedCloudBalancer and simulation of a federated cloud platform were carried out on CloudAnalyst Simulation toolkit. Experimental evaluations were carried out to check the effect of inter-cloud load balancer on the overall response time of the system and the overall processing time of the federated cloud environment. The results shows that the FedCloudBalancer with ACO performs well with an average response time of 92.33 ms as compared with 328.4 ms and 176.55 ms for Closest Datacenter (CDC) and Optimize Response Time (ORT) respectively. The FedCloudBalancer algorithm provides an improvement over the existing CDC and ORT inter-cloud load balancing algorithms using metaheuristic optimization approach